BACKGROUND: It was hypothesized that calcineurin inhibitors suppress vascular cyclooxygenase (COX)-2 and exert a reciprocal influence on in vivo prostacyclin and thromboxane. This could contribute to cardiovascular morbidity in transplanted patients. METHODS: The ability of immunosuppressives to suppress vascular COX-2 expression in vitro was studied in cultured human vascular smooth muscle cells. Blood and urine samples were collected from 28 renal transplant patients before and 2, 4 and 6 h after intake of immunosuppressives and from 11 controls. ELISA was used to measure (1) plasma 6-keto-PGF1alpha and TxB2; (2) urine excretion of PGI-M and TxB(2); (3) 6-keto-PGF1alpha in the whole-blood COX-2 assay; and (4) TxB2 in the whole-blood COX-1 assay. Platelet aggregation was measured optically. RESULTS: COX-2 in cultured vascular smooth muscle cells was suppressed by cyclosporine A (CsA); tacrolimus and rapamycin had no effect. Human renal arteries and vascular smooth muscle expressed calcineurin Abeta and Agamma isoforms. CsA had no effect on plasma 6-keto-PGF1alpha, whole-blood COX-2 activity or PGI-M urine excretion; after rapamycin intake, the former two increased. Plasma TxB2 did not change after drug intake. TxB2 in the COX-1 assay and urine excretion of TxB2 was significantly lower in tacrolimus- and rapamycin-treated patients compared to the CsA group. Platelet aggregation was increased significantly in the CsA group. CONCLUSIONS: Although CsA suppressed COX-2 in cultured vascular smooth muscle cells, systemic prostacyclin was not suppressed by either CsA or tacrolimus in vivo. Rapamycin and tacrolimus may actively suppress platelet and renal thromboxane formation. Differential changes in prostanoids may have implications for long-term cardiovascular hazard in patients treated with immunosuppressives.